Table 3 CNN architecture with its layers and parameters
Layer type | CNN-1 | CNN-2 | CNN-3 | CNN-4 |
|---|---|---|---|---|
Input | 109×1 | 109×1 | 109×1 | 109×1 |
First convolution | 16, 3×1, 2×1 | 16, 3×1, 2×2 | 16, 3×1, 2×1 | 16, 3×1, 2×1 |
Second convolution | - | 32, 3×1, 2×2 | 32, 3×1, 2×1 | 32, 3×1, 2×1 |
Third convolution | - | - | 64, 3×1, 2×1 | 64, 3×1, 2×1 |
Fourth convolution | - | - | - | 128, 3×1, 2×1 |
Fully-connected | 3×128 | 3×128 | 3×128 | 3×128 |
Output | 3×1 | 3×1 | 3×1 | 3×1 |